Manejo y visualización de datos

library(dplyr)

# El propio de dplyr
iris %>%
  filter(Sepal.Length > 5) %>%
  summarise(promedio = mean(Sepal.Width))

# (control  (command) + shift + M)

# el de R desde hace dos años
iris |>
  filter(Sepal.Length > 5) |>
  summarise(promedio = mean(Sepal.Width))

# 1. Cargar librerías

library(dplyr)
library(ggplot2)
library(readr)
library(palmerpenguins)
library(scico)

# 2. Leer datos
data(package = 'palmerpenguins')
datos <- penguins

# 3. Explorar

head(datos)

str(datos)

summary(datos)

unique(datos$island)

datos_filtrados <- datos %>% filter (year == 2007)

datos_filtrados <-datos %>% filter (year == 2007, sex=="female")

datos_filtrados <- datos %>% filter (year == 2007, sex=="female", island %in% c("Dream", "Bicoe")) 
# island == c("Dream", "Bicoe")) NO VALE!
# summarize
head (datos_filtrados, 3)

# 4. Limpiar: eliminar filas con NA
datos_limpios <- datos %>% 
  filter(!is.na(body_mass_g))

head(datos_limpios, 4)

# 5. Transformar: crear nueva variable
datos_limpios_kg <- datos_limpios %>%
  mutate(body_mass_kg = body_mass_g / 1000)

library(here)
library(tidyverse)
library(tm)
library(wordcloud2)
library(readxl)

# se puede importar tablas desde excel, csv, ggsheets, txt...

datos_curso = read_excel(here("data/datos-curso.xlsx"))
# ifselse en dplyr?
# Vemos un resumen
str (datos_curso)

#View (datos_curso) # Se nos abre en Rstudio

colnames(datos_curso) # Miramos las columnas que tiene

datos_curso_filtrado <- datos_curso |>
  rename (curso = "Curso de la primera matrícula en el programa de doctorado",
         experiencia = "Cuéntanos tu experiencia con R",
         comentarios = "Observaciones",
         email = "Correo electrónico") |>
  select (curso, experiencia, comentarios, email) |>
  mutate (email = str_replace(email, "anonymous", "anónimo")) |>
  filter (curso == "2022-23") # filtrado y sin filtrar otra diapo.
  
words <- tolower(
                unlist(
                  strsplit(datos_curso_filtrado$experiencia, " ") ) ) 

words <- gsub("[[:punct:][:digit:]]", "",words)
stop_es <- stopwords("spanish")
words_clean <- words[!words %in% stop_es]

# Plot
wordcloud2(data=table(words_clean), size=0.5,
           color = rep(RColorBrewer::brewer.pal(11, "PRGn"), 20))

plot(x=rnorm(100), y=rnorm(100))

ggplot(data = datos_limpios_kg, aes(x = bill_length_mm, y = bill_depth_mm)) + geom_point()

library(ggplot2)

p <- ggplot(data = penguins, aes(x = bill_length_mm, y = bill_depth_mm)) +
geom_point(aes(color = body_mass_g), alpha = .6) +
  scico::scale_color_scico(palette = "bamako", direction = -1)+ 
## custom axes scaling
scale_x_continuous(breaks = 3:6 * 10, limits = c(30, 60)) +
scale_y_continuous(breaks = seq(12.5, 22.5, by = 2.5), limits = c(12.5, 22.5)) +
## custom labels
labs(
title = 'Bill Dimensions of Brush-Tailed Penguins (Pygoscelis)',
subtitle = 'A scatter plot of bill depth versus bill length.',
caption = 'data = Gorman, Williams & Fraser (2014) PLoS ONE, graph = Cédric Scherer',
x = 'Bill Length (mm)',
y = 'Bill Depth (mm)',
color = 'Body mass (g)'
) + theme_minimal () 
p

dat <- data.frame(
  meal = factor(c("Breakfast", "Lunch", "Dinner", "Dinner", "Lunch")),
  total_bill = c(4.89, 7.23, 15.34, 21.23, 11.32)
)

head(dat)

dat <- data.frame(
  meal = factor(c("Breakfast", "Lunch", "Dinner", "Dinner", "Lunch")),
  total_bill = c(4.89, 7.23, 15.34, 21.23, 11.32)
)

dat %>%
  ggplot(aes(x=meal)) +
  geom_bar() + theme_minimal()

library(ggplot2)
library(dplyr)

dat <- data.frame(
  meal = factor(c("Breakfast", "Lunch", "Dinner", "Dinner", "Lunch")),
  total_bill = c(4.89, 7.23, 15.34, 21.23, 11.32)
)

# Very basic bar graph
dat %>%
ggplot(aes(x=meal, y=total_bill)) +
    geom_bar(stat="identity")

dat %>%
ggplot(aes(x=meal, y=total_bill)) +
    geom_bar(stat="summary", fun = "mean")

dat %>%
ggplot(aes(x=meal, y=total_bill, fill=meal)) +
    geom_bar(stat="identity")

dat %>%
ggplot(aes(x=meal, y=total_bill, color=meal)) +
    geom_bar(stat="identity")

dat %>%
ggplot(aes(x=meal, y=total_bill, fill=meal)) +
    geom_bar(stat="identity") +
    guides(fill=FALSE)

dat %>%
ggplot(aes(x=meal, y=total_bill, fill=meal)) +
    geom_bar(stat="identity") +
    guides(fill=FALSE) +
    labs(title = "Precio medio por comida", 
       subtitle = "Ajustado para dos personas",
       x = "Tipo de comida",
       y = "Precio") + theme_minimal()

dat <- data.frame(
  diet = factor(c("Vegan", "Carnivore", "Carnivore","Carnivore","Vegan", "Vegan", "Vegan")),
  meal = factor(c("Breakfast", "Dinner", "Breakfast","Lunch", "Dinner", "Dinner", "Lunch")),
  total_bill = c(5.32, 19.22, 4.89, 7.23, 15.34, 21.23, 11.32))

# Stacked bar graph -- queremos esto?
dat %>% 
  ggplot(aes(x=meal, y=total_bill, fill=diet)) +
  geom_bar(stat="identity")

dat %>%
ggplot(aes(x=meal, y=total_bill, fill=diet)) +
    geom_bar(stat="identity", position=position_dodge()) 

datos_limpios_kg %>% 
ggplot (aes(x = island, y = bill_length_mm)) + geom_boxplot()

violin_plot <- datos_limpios_kg %>% 
  ggplot (aes(x = island, y = bill_length_mm, fill = island)) + 
  geom_jitter (size=3, alpha=0.1, fill="grey") +
  #geom_jitter (aes(fill=island), size=5, alpha=0.2, fill="grey") +
  geom_violin(alpha=0.8) +  guides (fill= FALSE)+
  scale_fill_manual(values=c("#72874EFF", "#FED789FF", "#023743FF")) +
  theme_light() + theme(rect = element_rect(fill = "transparent"))

violin_plot

ggsave("images/violin_plot.png", violin_plot, width = 20, height = 20, units = "cm", scale = .5)

ggplot(datos_limpios_kg, 
       aes(x = flipper_length_mm)) + 
  geom_histogram() # bins=5

ggplot(datos_limpios_kg, 
       aes(x = flipper_length_mm)) + geom_density(bw=1)

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm)) + 
  geom_point() 

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = island)) + 
  scale_colour_manual(values=c("#72874EFF", "#FED789FF", "#023743FF")) +
  geom_point() + theme_light() 

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = bill_length_mm)) + 
  geom_point() + theme_light() 

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = bill_length_mm)) + 
  geom_point() +
  scale_color_gradientn(colors = c("#00AFBB", "#E7B800", "#FC4E07"))+
  theme_light() 

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = bill_length_mm)) + 
  geom_point() +
  scale_color_viridis_c(option = "plasma") + 
  theme_light() 

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = bill_length_mm)) + 
  geom_point() +
  scale_color_distiller(palette = "Spectral") +  # otras: "RdYlBu", "YlGnBu", "PuOr", etc.
  theme_light() 

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm)) + 
  geom_point() + geom_smooth()

# Method = "loess"

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm)) + 
  geom_point() + theme_light() + geom_smooth(method = "lm")

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = island)) + 
  scale_colour_manual(values=c("#72874EFF", "#FED789FF", "#023743FF")) +
  geom_point() + theme_light() + geom_smooth(method = "lm")

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = island)) + 
  scale_colour_manual(values=c("#72874EFF", "#FED789FF", "#023743FF")) +
  geom_point() + theme_light() +  facet_wrap(~species, scales = "free")

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = island)) + 
  scale_colour_manual(values=c("#72874EFF", "#FED789FF", "#023743FF")) +
  geom_point() + theme_light() +  facet_wrap(species~island, scales = "free")

datos_limpios_kg %>% 
  ggplot (aes(x = flipper_length_mm, y = bill_length_mm, colour = island)) + 
  scale_colour_manual(values=c("#72874EFF", "#FED789FF", "#023743FF")) +
  geom_point() + theme_light() +  facet_grid(species~island, scales = "free")

library(rgbif)
library(ggplot2)
library(rnaturalearth)
library(sf)

datos <- occ_search(scientificName = "Lynx pardinus", limit = 100)
df <- datos$data

# Basic map
world <- ne_countries(scale = "medium", returnclass = "sf")

# Plot
ggplot() +
  geom_sf(data = world, fill = "grey90", color = "grey30") +
  geom_point(data = df, aes(x = decimalLongitude, y = decimalLatitude),
             color = "darkred", size = 2, alpha = 0.7) +
  coord_sf() +
  theme_minimal()

# Basic map
spain <- ne_countries(country=c("spain", "portugal"))
ggplot() +
  geom_sf(data = spain, fill = "grey99", color = "grey30") +
  geom_point(data = df, aes(x = decimalLongitude, y = decimalLatitude),
             color = "darkgreen", size = 2, alpha = 0.7) +
  coord_sf() +
  theme_minimal()

library(FloraIberica)

map_distribution(genus = "Lavandula", species = "stoechas", size = 0.9)